Finding corresponding objects when integrating several geo-spatial datasets

Abstract

When integrating geo-spatial datasets, a join algorithm is used for finding sets of corresponding objects (i.e., objects that represent the same real-world entity). Algorithms for joining two datasets were studied in the past. This paper investigates integration of three datasets and proposes methods that can be easily generalized to any number of datasets. Two approaches that use only locations of objects are presented and compared. In one approach, a join algorithm for two datasets is applied sequentially. In the second approach, all the integrated datasets are processed simultaneously. For the two approaches, join algorithms are given and their performances, in terms of recall and precision, are compared. The algorithms are designed to perform well even when locations are imprecise and each dataset represents only some of the real-world entities. Results of extensive experiments show that one of the algorithms has the best (or close to the best) performances under all circumstances. This algorithm has a much better performance than applying sequentially the one-sided nearest-neighbor join.

Entropy-Weighted Instance Matching Between Different Sourcing Points of Interest

“…In the comparison of the ontology approach, it does not need to construct a complicated domain ontology library or repository of knowledge, which appears to be more practical for commercial POI providers. Beeri [4] and Safra, et al [7] assumed that associated objects are closer to each other in the spatial attribute. Based on this assumption, they argued for an algorithm that has higher accuracy than the unilateral nearest neighbor algorithm and described the algorithm in parallel and/or series forms.…”

“…In the GIS field, the selection of the training dataset is always performed manually [1,4,12,21,32]. Similarly, we select 300 POIs in Ds randomly and then find 253 matching POIs in Db manually; the pairs of POIs are called dataset Dm.…”

“…However, POIs from different sources (without a global identifier) [4] with diversity not only in data structures, content, emphases and coverage area but also in the properties (e.g., category, address) However, POIs from different sources (without a global identifier) [4] with diversity not only in data structures, content, emphases and coverage area but also in the properties (e.g., category, address) that contain semantic features exhibit discrepancy to some degree. In addition to the problem of inconsistencies caused by the shortage of global identifiers, even the matched attributes have significant discrepancies.…”

Pattern Classification Approaches to Matching Building Polygons at Multiple Scales

“…In general, those measures can be divided into geometric, semantic and contextual measures. For instance, Beeri et al (2005) developed spatial join algorithms that match points only using their locations. To match more complex objects (polygons and networks), other geometric information such as angles, shapes, topological properties are also used (Walter and Fritsch, 1999;Gösseln and Sester, 2004).…”

Computing a k-Route over Uncertain Geographical Data

“…A singleton (i.e., a set that contains a single object) is correct if it represents a real-world entity that does not have a corresponding object in the other source. In the absence of keys, integration can be done by using object locations [3,4] or by using locations and additional attributes [17]. However, since locations are inaccurate, it is uncertain whether any given pair of the result is correct, that is, whether its two objects indeed represent the same real-world entity.…”